Bug 162579 - Add an Option to disable the CodeCache
Summary: Add an Option to disable the CodeCache
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: JavaScriptCore (show other bugs)
Version: WebKit Nightly Build
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Joseph Pecoraro
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2016-09-26 15:06 PDT by Joseph Pecoraro
Modified: 2016-09-26 19:57 PDT (History)
6 users (show)

See Also:


Attachments
[PATCH] Proposed Fix (3.03 KB, patch)
2016-09-26 15:09 PDT, Joseph Pecoraro
no flags Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Joseph Pecoraro 2016-09-26 15:06:58 PDT
Summary:
Add an Option to disable the CodeCache

Useful when benchmarking to ensure we do more parsing.
Comment 1 Joseph Pecoraro 2016-09-26 15:09:50 PDT
Created attachment 289879 [details]
[PATCH] Proposed Fix
Comment 2 Joseph Pecoraro 2016-09-26 15:10:29 PDT
Comment on attachment 289879 [details]
[PATCH] Proposed Fix

View in context: https://bugs.webkit.org/attachment.cgi?id=289879&action=review

> Source/JavaScriptCore/ChangeLog:8
> +        * runtime/CodeCache.cpp:

We could also use this in EvalCodeCache. Let me know if you think we should as well.
Comment 3 Geoffrey Garen 2016-09-26 15:17:07 PDT
Comment on attachment 289879 [details]
[PATCH] Proposed Fix

r=me
Comment 4 WebKit Commit Bot 2016-09-26 15:50:38 PDT
Comment on attachment 289879 [details]
[PATCH] Proposed Fix

Clearing flags on attachment: 289879

Committed r206400: <http://trac.webkit.org/changeset/206400>
Comment 5 WebKit Commit Bot 2016-09-26 15:50:43 PDT
All reviewed patches have been landed.  Closing bug.
Comment 6 Saam Barati 2016-09-26 19:56:12 PDT
Comment on attachment 289879 [details]
[PATCH] Proposed Fix

View in context: https://bugs.webkit.org/attachment.cgi?id=289879&action=review

> Source/JavaScriptCore/runtime/Options.h:391
> +    v(bool, useCodeCache, true, Normal, "If false, the parser will not use the code cache.") \

Nit: This might be confusing with the word "parser" in it since the parser has its own cache and this cache is on top of the parser. I'd just say: "If false, we won't use the unlinked byte code cache" or something similar.
Comment 7 Saam Barati 2016-09-26 19:57:31 PDT
(In reply to comment #2)
> Comment on attachment 289879 [details]
> [PATCH] Proposed Fix
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=289879&action=review
> 
> > Source/JavaScriptCore/ChangeLog:8
> > +        * runtime/CodeCache.cpp:
> 
> We could also use this in EvalCodeCache. Let me know if you think we should
> as well.

They're two different caches. I'd say that if we find an option for the eval code cache useful we can add it in a separate patch. I think it's good to have this option just be for the unlinked code cache.